This is a talk given at the Redwood Center for Theoretical Neuroscience, UC Berkeley on May 8, 2007. Speaker is Lokendra Shastri from ICSI (International Computer Science Institute) and UC Berkeley. http://www.icsi.berkeley.edu/~shastri/.
We readily remember events and situations in our daily lives and acquire memories of specific facts by watching a telecast or reading a newspaper. This one-shot mnemonic ability poses a challenge for computational neuroscience: How does the brain carry out this remarkable memorization task so rapidly and effortlessly? There is a broad consensus that the hippocampal system (HS), consisting of the hippocampal formation and neighboring cortical areas plays a critical role in encoding and retrieving such "episodic" memories. Furthermore, a great deal is known about the anatomy, local circuitry, and plasticity of the HS, including quantitative data about cell counts and connectivity. But how the HS supports episodic memory function, and what constitutes an "episodic memory trace" is not well understood. I will describe a circuit-level computational model (SMRITI*) that demonstrates how a cortically expressed pattern of rhythmic activity representing an event can be transformed rapidly (~500 msec.) into a robust and functionally adequate memory trace in the HS. This memory trace consists of a complex neuronal circuit having several functional components. A quantitative analysis of the model indicates that memory traces are sparse and specific but, at the same time, they are highly redundant and physically dispersed. The model suggests a striking correspondence between the functional requirements of encoding episodic memories and the idiosyncratic architecture of the HS. Recent empirical findings show that the match between the model and the HS extends to the level of detailed micro-circuits identified in the HS. Time remaining, I will also discuss some of the model's predictions about the nature of memory deficits resulting from insult to various regions and pathways of the HS.